DataBase as Documentary

How can NHS databases be a vehicle for mass participation art projects that critically encourage well-being?

Graham Harwood, January 16th 2011

Introductory notes:

This research cuts across the discourses of Computer Science, Public Health, Art and their methods and modes of academic research. I'm not an expert in any of these areas but have enough curiosity to weave together various debates, threads, critiques that I have come across as I have drifted over the terrain. I have drawn together this introduction in an attempt to allow for a general preamble to the area aimed at allowing us to work together in the future.

John Grierson defined the term documentary in the 1920's before making films for the General Post Office(GPO). He dubbed this new genre a“creative treatment of reality”. His interest, like that of this research, was in a socially active aesthetics. Since the 1930's we have witnessed documentary's ability to influence public perception. Examples range from Leni Riefenstahl’s fascist work ‘Triumph of the Will’ (1935) aestheticising politics and ascribing the medium’s power to a particular personality, to current work on the Afghanistan conflict trying to critique the political status quo. The energies contained within the documentary can empower governance and its mirrors in a multitude of ways but at the same time they may make an interesting, innovative and popular art possible.

Maternity: a film of Queen Charlotte's Hospital, England, Personal Films, 1935 addressing the then fact that in Britain every two hours a women dies in childbirth

This research adopts the view that databases carry the same seeds of creativity that early documentary makers saw in film. Both have a clear relationship to a form of factual truth that can empower, both have the capacity for propaganda and deception, both form views which are privileged by their authors. The film director directs the eyepiece of the camera through the conduct of the cameraman and the editor shapes and cuts the film to create an agreed view for a set audience and the promotion of a particular ideology. The database administrator also constructs views of the data that can only be seen by particular roles within the enterprise. In association with analysts they also produce queries that iterate over the data atoms to configure them into particular narratives formed from the enterprise’s ideology.

Early documentary was at it's best when it put its subject at the centre of the film, allowing the subject to express an intelligent voice about the social and material relations in which they found themselves. People were filmed demonstrating pride in their skills or bravery in the face of adversity such as in Granton Trawler (1934) in which Grierson lives aboard the Isabella Greig. He becomes part of its community and shares its problems, exposing the working conditions on board the trawler as the men fish, haul, gut and sort the catch as it carries out its dragnet fishing along the Viking Bank off the Norwegian coast. When documentary fails as in The Fairy of the Phone (1936) it was due to it taking a disciplinary stance in which the audience is expected to regulate their behaviour in accordance with the norms shown in the film. This form of documentary, the ‘public information film’ has quite rightly been subjected to much disdain and satire.

In the 21st Century, governments and other enterprises are increasingly trying to convince us of the usefulness of our personal data for the public good. (With the proviso that it is efficiently collected, quality controlled and managed securely.) This process, we are informed, will allow us to reap the various forms of value/benefit produced by improving myriad forms of efficiency. If we think about government as a series of tactics, strategies, techniques, programmes and aspirations of those authorities who wish to control, influence or improve what we think of and do as a population, databases inform various modes of thinking, decision making and acting. We hope that the information we give up will be used in a timely manner and hopefully help them make the best use of public finances and to evidence decisions in a rational and reasonable manner. We also have the expectation that this process will be as open and transparent to the public as possible while protecting our individual privacy.

Michel Foucault – Security, Territory, Population -isbn 978-1-4039-8633-5
General features of apparatus of scurity -p55 - 86

In it's attempt to convince us about parting with or not complaining to much about the use of our personal data, government has set up a number of open data initiatives. A recent high-profile case would be the publishing of MP's allowances. The website mpsallowances.parliament.uk allows anyone with an internet connection to view all the MP's expenses in something resembling a software panoptican, an equal gaze over all the MP’s allowances in which anomalous spending of the bad may be foregrounded against the normal spending of the good.

The use of this form of equal gaze of many eyes has steadily been heading up the political armoury for the last decade as computing power, and databases in particular, have inflated new forms of authority by creating new views of information that have been processed into new forms of knowledge through relational machines. Many initiatives have arisen over the past decade around this opening up of data but one which is particularly illustrative for this report is the development of the powerful Dr Foster Intelligence Ltd.

Health surveillance, Panopticons and more equal gaze

Databases as a tool of health surveillance attempts to establish an equal gaze which we can think of here as a technical process of ordering population data so that the information can be queried. Anomalies become triggers for new tactics, strategies, techniques, programmes or for checking the aspirations/targets of government. This system has emerged from the concerns of what Foucault called the "social medicine" of the early 19th Century, born of the miasma theory of disease and urban concerns for the management of populations, from which Foucault set out his concept of "biopower".

16 years ago, Professor Brian Jarman joined others in creating The Hospital Standardised Mortality Ratio which revealed, among other things, that even when controlled for variations like patients age or class, English mortality rates differed up to 76 per cent between area to area.

Dr Foster has been publishing this mortality data since 2000. The HSMR is a calculation used to monitor death rates in a trust and is based on a subset of diagnoses which give rise to 80% of in-hospital deaths. HSMRs are based on the routinely collected administrative data often known as Hospital Episode Statistics (HES), Secondary Uses Service Data (SUS).

In 2009 Dr Foster Intelligence revealed that Basildon and Thurrock University Hospitals NHS Foundation Trust had a mortality rate a third higher than the national average: about 350 more people died in a year more than would have been expected, leading to enquiries that revealed the appalling state of Basildon Hospital.

So good quality databases can save lives or at least find out where they are being needlessly lost. When databases find a critically responsive audience they introduce new forces into the technologies of power. Tim Kelsey, a co-founder of Dr Foster Intelligence, wrote an article for Prospect Magazine in August 2009 arguing for 'a smarter use of public service statistics that can save lives as well as money'. He argues that this vision is threatened by 'anxious civil libertarians' who want to stop the state sharing our personal data.

Tim Berners-Lee, founder of the World Wide Web, has also been a keen advocate of the Open Data Movement of the type that revealed the appalling conditions at Basildon Hospital. He often sites the data for cycling accidents that was released by the UK and was transformed into a map by the timesonline.co.uk two days after release.

http://labs.timesonline.co.uk/2009/cycling_accidents/

The map is credited with putting pressure on local authorities to clear up accident hotspots and enabling cyclists to become aware of potentially hazardous areas.

The UK Government itself sees releasing public data as a way to help people understand how government works and how policies are made – there is move towards an equal gaze for everyone in which we all police each other, as in the case of MP's expenditure. The HM treasury has estimated that open data is worth £6billion to the economy with data.gov.uk bringing all its central data together in one searchable website. It believes that making this data easily available will make it easier for people to make decisions and put forward suggestions for government policies based on detailed information. (www.data.gov.uk)

Within the ambition of these forms of open data initiatives especially those of government, problems arise when we begin to consider the prerequisites for participation which fall technically within the area of transparency. The Open Data Foundation (ODF) suggests the following guide for open data (www.opendatafoundation.org)

For ODF, transparency means being to:

Discover the existence of data
Access the data for research and analysis
Find detailed information describing the data and its production processes
Access the data sources and collection instruments from which and with which the data was collected, compiled, and aggregated
Effectively communicate with the agencies involved in the production, storage, distribution of the data
Share knowledge with other users

Government data produced under this notion of transparency can be viewed as operating the ventricles of an enlightened power, interconnecting the domains of government and population. The relative openness of the data can be seen as an attempt to unfold ‘rationalist’ attempts to evidence decisions. This transparency debate creates a protocol between government and non government Data Base Management System administrators and ethical statistical analysts who summon the latent energies contained in the new knowledge to power there differing political factions. This is a data exchange between those who can already perceive data from it's modes of representation or to put it another way understand the construction of the data and wish to exploit it as a form of self reflexive critic of government.

As with Doctor Fosters, much of this debate crosses over into other forms of discourse such as healthcare and security or economy. Due to historical and social formations too numerous to mention in this report, the gap between the wider public's perception of data and the social experience it attempts to model, creates a form of indifference toward the expectations of this kind of initiative. A partial remedy for this indifference might be found in making data more vital through taking a more critical view of transparency. This would require seeing it, not so much in technical terms – the protocols of the enlightened yet unequal participants of the governed and government - but more in terms of the data itself having some kind of agency.

Following the Agency of Data

Such a perspective can be imagined through a critical reading in which we are able to see what decisions the data has informed and evidenced and how that data has been collected, for what purpose and by whom. Taking this thread a little further It would also be illuminating to see what positions the data places the subject of its records, and where too it places the user of the data.

While this form of analysis is not possible or interesting for everyone, academic institutions such as the Centre of Cultural Studies Goldsmiths have a good track record of critical thinking in this area. Our research at Goldsmiths would probably place something like the data.gov.uk within the domain of technologies of power or to put it another way, amongst the relationship between knowledge and power.

Within the technologies of power, the database can be seen as an energy source, a motor of change or an amplifier for the progression of truths within the discourses that fabricate them. 'Truth', in this instance, should be understood as the system of ordered procedures for the production, regulation, distribution, circulation and operation of authoritative statements. We can think of The Hospital Standardised Mortality Ratio as just such a truth-creating machine that would be thought of as penetrating illusions and seeing through to underlying realities. We do not have to agree with the truth of these machines as we are not trying to say they are true to everyone, only true to the discourses that produce them.

At it's most simplistic the relational machine, (the conceptual machine that make a database possible) operates as a process where data atoms are placed in entities and relations, queries then process those atoms into information. New knowledge is formed by comparing the information. Power then emerges as the new knowledge which has the potential to change the conduct of others.

E.g The dataset of bicycle accidents is published and by comparing it's information ( through data mash-ups) the energy contained in it forms new knowledge through the map which changes the conduct of the cyclists.

A somewhat similar system can be thought to operate in health discourses through the formation of expert knowledge and the power that accompanies it. This power potentially offers the discourse that produced it, control at a distance. This form of health promotion as a form of government produces modern subjects: it defines empirically what it is to be healthy and it ‘supervises’ the proper routes to health through a discipline. The problem with scientific knowledge translated into appropriate behaviours for individuals to follow is that it requires a self-reflective, self-regulating individual with the correct concern for themselves.

Many people however are dis-interested in the proper routes to health – they see this form of discipline as having very little gravity in their lives compared with the circumstances they find themselves in and would rather rely on their own intelligence and that of their communities, they refuse to be the subject of this control or discipline. The credibility of the data model is called into question by it's abstraction from the social experience that formed it.

The collecting of personal information for evidencing decisions is now at the core of 21st Century Politics, Security, Economy, Health and Democracy yet work is only beginning to find out how the relational machine (databases) affects change around them and how they transform those enterprises that undertake them.

In the “Joint Strategic Needs Assessment” Developed by the Liverpool NHS Primary Care Trust (PCT) dated 2008. The PCT had found that it had a strong notion that 10,000 people were out there somewhere with Hypertension and that if they could be found then morbidity rates through out the city could be reduced and this might boost inward investment in the city. This speculation was created by comparative analyses involving many datasets that set up forms of authority which could direct the PCT priorities. From talking to people in the PCT it became obvious that people thought that much better value could be made of the money if spent on 16-24 year old's but this would not be evidenced within the life span of a parliament or two and therefore would not fit well into health governance.

On a recent trip to Bristol working on ‘B-Open’, another open data initiative, I was shown around the basement of the City’s Council House. Large open-plan offices set in the now vaporised finance area of the council in which people used to pay their rents and rates. The rooms were as dull as to be expected in the seat of power but at each entrance to a subsection was a card reader. We constantly had to bother someone to move from one regulated area to another. When I asked what was at the core of this regulation I was told that it was in readiness for the introduction of ContactPoint a database established by the previous Labour administration to improve child protection after the Victoria Climbie child abuse scandal. ContactPoint was to hold the records of over 11 million children in danger in England was scrapped by the new Conservative and Liberal Democrat government.

Peoples conduct was being ordered by the potential knowledge/power relation of non-existent machine. During this phase of my database research people constantly slip me other such anecdotes on the agency of databases. Whether when filling them up or querying them, or viewing their records, people recount how their lives are effected and how this contemporary machine transforms the enterprises for whom they work.

If we are to think about how NHS databases can be be a vehicle for mass participation art projects that encourage reflection on well-being. Then the project must engage critically with the gap between the wider public's perception of data and the social experience it attempts to model and also find a way to engage people who are dis-interested in the proper routes to health defined by the disciplinary method described above.

To further our speculation of how this might be achieved would require mapping the various meanings and agencies arising from databases and the relational model. This would be a very large area of research and beyond the scope of this report. In a humble and clumsy way I have tried to sketch out some possible methodologies to examine these issues by exploring the Public Health Birth Record file from Liverpool Primary Care Trust and try to envision how it might be used.